Mathematics · Point-Set Topology

Topology
A Visual Introduction

The study of spaces, continuity, and the properties that survive deformation.

What is a Topology?

Topology abstracts the intuitive notion of "nearness" without relying on distance. It is the skeleton of continuity itself.

A topology on a set X is a collection τ (tau) of subsets of X, called open sets, that satisfies three fundamental axioms. The pair (X, τ) is called a topological space.

Unlike metric spaces, topology doesn't care about distances — only about which sets are declared "open." This makes topology extraordinarily general: the same framework describes the real line, a donut, a sphere, and stranger objects alike.

Core Intuition: Think of open sets as neighborhoods of "fuzzy" regions — every point in an open set has some wiggle room. Topology axiomatizes exactly what constraints these regions must satisfy to behave coherently.

The Basic Setup

Let X be any non-empty set — the "universe" of points. A topology τ on X is a family of subsets satisfying the three axioms on the next tab. We call the elements of τ the open sets of the topology.

A set X with a collection of subsets declared as open. The open sets must include ∅ and X, and be closed under unions and finite intersections.

The Three Axioms

Every topology — no matter how exotic — must obey exactly these three rules. They are the constitution of topological space.
01
Trivial Sets

Both the empty set and the entire space must be open sets in every topology.

∅ ∈ τ and X ∈ τ
02
Arbitrary Unions

The union of any collection of open sets — even infinitely many — must also be an open set.

If Uᵢ ∈ τ for all i,
then ⋃ Uᵢ ∈ τ
03
Finite Intersections

The intersection of any finite collection of open sets must also be open. (Infinite intersections are not required to be open.)

If U₁, ..., Uₙ ∈ τ,
then U₁ ∩ ... ∩ Uₙ ∈ τ

Why finite intersections only?

Infinite intersections of open intervals can collapse to a single point, which is closed — not open. For example, in ℝ with the standard topology:

⋂ₙ₌₁^∞ (−1/n, 1/n) = {0}

The intersection of infinitely many open intervals narrows down to the single point {0}, which is not open in the standard topology on ℝ. This is why Axiom 3 requires only finite intersections.

Closed Sets are defined as complements of open sets. A set can be: open, closed, both (clopen), or neither. The terms are not opposites!

Visual demonstration of Axioms 2 and 3: two open sets U and V (left), their union U∪V (middle, must be open), and their intersection U∩V (right, must be open for finite collections).

Simple Topologies

The simplest topologies are defined on small finite sets. They reveal the axioms in their purest form.

Let X = {a, b, c}. How many distinct topologies can we define on X? (The answer is 29.) Here are the most illuminating ones:

Name Open Sets τ Complexity Notes
Indiscrete {∅, X} simple Minimal. No points can be separated.
Discrete {∅,{a},{b},{c},{a,b},{a,c},{b,c},X} simple Maximal. Every subset is open.
Particular Point {∅, {a}, {a,b}, {a,c}, X} medium Every open set must contain point a.
Excluded Point {∅, {b}, {c}, {b,c}, X} medium Every proper open set excludes point a.
Sierpiński Space {∅, {1}, {0,1}} simple X = {0,1}. The simplest non-trivial T₀ space.

Indiscrete Topology

Only ∅ and X are open. Points are completely inseparable — no open set can distinguish between them.

Discrete Topology

Every subset is open. Points are maximally separated — every singleton {a}, {b}, {c} is its own open set.

Particular Point (a)

Every non-empty open set must contain point a. It acts as a "mandatory member" of all open neighborhoods.

Sierpiński Space

X = {0,1}. The open sets are ∅, {1}, and {0,1}. The simplest space that is T₀ but not T₁.

Complex Examples

Moving to infinite sets and geometric spaces, topologies become far richer — and far more interesting.

Standard Topology on ℝ

The most important topology in analysis. A subset U ⊆ ℝ is open if every point in U has an open interval around it entirely contained in U. Open sets are arbitrary unions of open intervals (a,b).

Open Intervals on ℝ

Open sets in ℝ are unions of open intervals. Endpoints are excluded. Any union of such sets is also open.

Cofinite Topology on ℝ

A set is open iff its complement is finite (or it's ∅). This is coarser than the standard topology.


Product Topology

Given two topological spaces (X, τ_X) and (Y, τ_Y), the product topology on X × Y has as basis all sets of the form U × V where U is open in X and V is open in Y. The plane ℝ² with its standard topology is the product of ℝ with itself.

Product Topology — ℝ × ℝ = ℝ²

Open sets in ℝ² are unions of open rectangles (U × V). A disk is also open — as a union of infinitely many open rectangles.


Quotient Topology

Given a space X and an equivalence relation ~, the quotient space X/~ identifies equivalent points. The quotient topology makes the projection map q: X → X/~ continuous. Gluing the ends of an interval produces a circle; gluing opposite edges of a square produces a torus.

Interval → Circle

Identify endpoints 0 ~ 1 in [0,1]. The quotient space is homeomorphic to S¹.

Square → Torus

Identify top↔bottom and left↔right edges of [0,1]². The quotient is the torus T² = S¹ × S¹.


Subspace Topology

Given (X, τ) and a subset A ⊆ X, the subspace topology on A is τ_A = {U ∩ A : U ∈ τ}. The open sets of A are exactly the intersections of open sets of X with A.

Subspace of ℝ²

The interval (0,1) is open in ℝ. Its intersection with [0,1] is (0,1) — still open in the subspace. But {0} ∩ (−ε, ε) = {0} is open in the subspace {0,1} of ℝ.

Key Concepts

Armed with a topology, we can define continuity, convergence, compactness, and connectedness — all without a ruler.
Hausdorff (T₂)

Any two distinct points can be separated by disjoint open sets. Most "reasonable" spaces are Hausdorff.

∀ x≠y ∃ U∋x, V∋y : U∩V=∅
Connectedness

A space is connected if it cannot be split into two disjoint non-empty open sets. ℝ is connected; {0,1} (discrete) is not.

X ≠ U ⊔ V for any open U,V
Compactness

Every open cover has a finite subcover. Compact sets generalize "closed and bounded" from ℝⁿ to arbitrary spaces.

X = ⋃Uᵢ ⇒ X = ⋃ᵢ₌₁ⁿ Uᵢ
Continuity

f: X→Y is continuous if the preimage of every open set in Y is open in X. This replaces the ε-δ definition.

V open in Y ⇒ f⁻¹(V) open in X
Homeomorphism

A bijection f: X→Y where both f and f⁻¹ are continuous. Homeomorphic spaces are "topologically identical."

f, f⁻¹ continuous bijections
Basis

A collection ℬ of open sets such that every open set is a union of basis elements. A basis generates the topology.

τ = {⋃ B : B ⊆ ℬ}

Connectedness

Connected vs disconnected spaces. A space is disconnected if it decomposes into two disjoint clopen sets.

Compactness — Open Cover

[0,1] is compact: any open cover reduces to finitely many sets. (0,1) is not compact: the cover (1/n,1) has no finite subcover.

The Topologist's Famous Joke: A topologist is someone who cannot distinguish between a coffee mug and a donut — both are genus-1 surfaces, homeomorphic to each other. Topology studies what survives continuous deformation.